Back to EveryPatent.com
United States Patent |
5,682,037
|
de Cesare
,   et al.
|
October 28, 1997
|
Thin film detector of ultraviolet radiation, with high spectral
selectivity option
Abstract
Thin film detector of ultraviolet radiation with high spectral selectivity
option, and a structure placed between two electrodes, formed by the
superposition of semiconductor thin films such as hydrogenated amorphous
silicon and its alloys with carbon. The device is able to absorb a large
quantity of UV radiation and to convert it into electric current being
transparent to photons of longer wavelengths. Its deposition technique
allows fabrication on substrates of glass, plastic, metal, ceramic types
of materials (also opaque, also flexible), on which a conductor material
film has been predeposited. It can be fabricated on substrates of any
size.
Inventors:
|
de Cesare; Giampiero (Rome, IT);
Irrera; Fernanda (Rome, IT);
Palma; Fabrizio (Rome, IT)
|
Assignee:
|
Universita Degli Studi Di Roma "La Sapienza" (Rome, IT)
|
Appl. No.:
|
588110 |
Filed:
|
January 18, 1996 |
Foreign Application Priority Data
| Feb 09, 1995[IT] | RM95A0073 |
Current U.S. Class: |
250/372; 250/370.01; 257/53; 257/55; 257/56; 257/62; 257/63; 257/77; 257/458; 257/E31.062 |
Intern'l Class: |
H01L 027/14 |
Field of Search: |
250/372,370.01,370.06
257/55,56,62,63,458,53,77
|
References Cited
U.S. Patent Documents
4177474 | Dec., 1979 | Ovshinsky | 257/55.
|
4217598 | Aug., 1980 | D'Auria et al. | 257/458.
|
4329699 | May., 1982 | Ishihara et al. | 257/55.
|
4469715 | Sep., 1984 | Madan.
| |
4772933 | Sep., 1988 | Schade | 257/458.
|
4839240 | Jun., 1989 | Shimizu et al.
| |
4857115 | Aug., 1989 | Iwamoto et al. | 257/53.
|
5015838 | May., 1991 | Yammagishi et al. | 257/458.
|
5073809 | Dec., 1991 | Bigan et al. | 257/458.
|
5097306 | Mar., 1992 | Tokuda | 257/458.
|
5210766 | May., 1993 | Winer et al. | 257/53.
|
5282993 | Feb., 1994 | Karg | 257/56.
|
5311047 | May., 1994 | Chang | 257/55.
|
5414275 | May., 1995 | Sugawa et al. | 257/458.
|
5557133 | Sep., 1996 | De Cesare et al. | 257/458.
|
Primary Examiner: Tokar; Michael J.
Assistant Examiner: Tyler; Virgil O.
Attorney, Agent or Firm: Dubno; Herbert
Claims
We claim:
1. A photodetector effective in the ultraviolet region of the spectrum,
comprising:
a substrate;
a first electrode on said substrate;
a second electrode spaced from said first electrode; and
a stack between said electrodes of at least three thin-film layers forming
a p.sup.+ n.sup.+ rectifying junction and including in succession a
p.sup.+ amorphous hydrogenated and carbon-containing silicon layer
photogenerating charge carriers upon illumination by light of said
ultraviolet region, an intrinsic amorphous hydrogenated and
carbon-containing silicon layer, and an n.sup.+ amorphous hydrogenated
silicon layer.
2. The photodetector defined in claim 1, further comprising an additional
stack of thin films of hydrogenated amorphous silicon or amorphous
hydrogenated and carbon-containing silicon layers responsive to a
frequency band in a portion of the spectrum other than said ultraviolet
region.
3. The photodetector defined in claim 2 wherein said frequency band has
greater wavelengths than those of said ultraviolet region.
4. The photodetector defined in claim 1 wherein said stack has a p.sup.+
-i-n.sup.+ -i-n.sup.+ -i-p.sup.+ structure.
Description
FIELD OF THE INVENTION
Our present invention relates to a sensor that permits detection of
ultraviolet light (UV. with .lambda.<400 nm with high rejection of
radiation of longer wavelengths.
SUMMARY OF THE INVENTION
The photodetector of the invention comprises a p.sup.+ -n.sup.+ junction of
hydrogenated amorphous silicon-carbon alloy (a-SiC:H) between two
electrodes connected with the exterior and which is fabricated by thin
film technology. The temperatures during the whole fabrication process are
such that the photodetector can be made on any substrate such as glass,
plastic and metal. Thanks to the technology used, fabrication is possible
in the form of large-area, high resolution two-dimensional matrixes.
The present invention, therefore, is in the field of UV radiation detection
and finds application in the detection of ultraviolet components of light
on an unwanted background of visible and/or infrared radiation which
should be filtered through in a case in which a high spatial resolution of
a two-dimensional image is necessary.
The invention can be used to detect phenomena accompanied by emission of
light: for example in instrumentation for detection of astronomic and
astrophysical data: in instrumentation for control of reactions in
chemical and biological fields; and in ionized gas spectroscopy (fusion
plasma and electric shock plasma: for clinical diagnostics, etc.).
It can be employed in the development of:
detection systems with one window functioning in the ultraviolet spectrum
and with one or two additional windows functioning in other spectral
regions;
ultraviolet radiation detection systems with high rejection of radiation at
long wavelengths;
large-area high resolution matrix systems capable of detection in the
ultraviolet spectral range.
The focal point of the invention is its capacity for absorbing a large
quantity of ultraviolet radiation and converting it into electric current,
letting the photons of other spectral ranges pass unhindered. In this
invention the UV light absorption is made possible by the following
factors:
the frontal electrode is a metal grid which offers to the incident light
open areas through which the light can pass;
the UV that passes through these open areas is absorbed in the p.sup.+
layer and converted into electric charge carriers; and
the electrons photogenerated in the p.sup.+ can reach the rear electrode,
crossing the entire structure.
Selectivity in the ultraviolet range is possible if the thickness of the
detector is much below the dimension at which there is a strong absorption
of the visible and infrared radiation.
BACKGROUND OF THE INVENTION
Currently some highly sensitive ultraviolet light photodetectors are
commercially available. These UV detectors can be divided into two
categories: (a) single-channel detectors, for a punctual (point) detection
and (b) multi-channel detectors for a two-dimensional (linear) detection.
The single-channel detectors include photomultiplier tubes and the
crystalline silicon photodiodes. Multi-channel detectors are: crystalline
silicon photodiode arrays, charge-coupled devices (CCD) and multi-channel
plates (MCP). The well-known photomultiplier tubes permit the detection of
ultraviolet radiation with high sensitivity and selectivity with respect
to the visible part of the spectrum by appropriately selecting the cathode
material. However, the use of photomultipliers has a number of
disadvantages which are overcome by the present invention. In fact,
photomultipliers require typical power supply voltages exceeding one
thousand volts. They are vacuum tubes, difficult to handle, bulky and do
not allow the integration of several elements.
Crystalline silicon photodiodes have an optimum efficiency in the visible
spectral range but can allow the detection of ultraviolet radiation only
after sophisticated and expensive mechanical and optical processing. They
require low power supply voltages and can be integrated in arrays of a few
centimeters in size.
The UV-CCD's are crystalline silicon components for which very special
processing is necessary as well. They are multi-channel detectors which
are highly sensitive with high signal/noise (S/N) ratios, especially if
they work at low temperatures. Basically CCD's are analog linear shift
registers. The electrons photogenerated within the silicon are collected
in a matrix of pixels which are then read sequentially. The
two-dimensional image can thus be reconstructed. There are at least three
major disadvantages in detection of ultraviolet radiation by CCD's namely,
the cost the impossibility to provide large-area two-dimensional matrixes
and the necessity to filter the visible radiation in the case that
ultraviolet components must be detected against background of other
radiation.
The Micro-Channel Plates (MCP) amplify even very weak light signals,
through a cascade multiplication process. They consist of millions of
microscopic conductive glass tubes used fused together in a disc-form
base, to the heads of which is applied a high potential difference
(typically of 1,000 V). The MCP's can function as image intensifiers or as
photon counters and, thus, have an excellent sensitivity. They are
provided at the input stage, with the deposition of a selective
sensitivity photocathode and at the output stage, with a phosphorous
shield. Due to their high sensitivity and S/N ratio performances, MCP's
are used in space applications and in astrophysics. However, the wide
commercialization of MCP's has encountered difficulty because of the high
cost of the MCP due to the technological complexity and need for the 1,000
V power supply.
ADVANTAGES OF THE INVENTION
With respect to the commercial UV photodetectors, the invention solves the
problems of filtering of the background visible and infrared radiation;
electric power consumption; large areas integration and, in addition, is
less costly.
The invention allows optimization of the thicknesses and coefficients of
absorption of the amorphous semiconductor layers composing the junction,
as well as of the geometry of the metal grid that serves as the front
electrode. Besides increasing the maximum detection efficiency in the UV,
this optimization allows also tuning of the operating band of the detector
by shifting it toward the near UV radiation or toward the far UV
radiation, depending on the applications. In fact, it had already been
demonstrated by other researchers that, by acting on the deposition
parameters and on the type and concentration of impurities in the silicon
alloy, absorption could be enhanced in the visible range toward the UV or
toward the infrared radiation. Thus the physical parameters to be
optimized are: a) the absorption profile and b) the thickness of the
detector. This optimization can be obtained by the control of the
deposition parameters, namely, the deposition time and the percentage of
carbon in the alloy.
The optimization and the reproducibility of the thicknesses of the layers
is made possible by control of the Glow Discharge time of deposition the
other parameters being fixed. The coefficients of absorption of the
hydrogenated amorphous silicon allow instead depend upon the fundamental
properties of the material such as, the extension of the energy and
optical gap of the semiconductor and the density of the defect states in
the gap. These, in turn, depend upon the growth parameters in a very
complicated way. A simple and repeatable method for varying the absorption
coefficient profile as a function of the wavelength is to form
silicon/carbon or silicon/germanium alloys in known percentages. This is
obtained through the introduction of a controlled flow, respectively, of
methane or germanium gas in the deposition chamber. The resultant
carbon/silicon alloy is an amorphous semiconductor of a greater energy gap
than the amorphous silicon, which penalizes the absorption of the visible
and of the infrared with respect to the ultraviolet radiation.
However, the a-SIC alloy to be used in the device must not contain too high
percentages of carbon with respect to silicon because its electronic
properties would prove poor.
BRIEF DESCRIPTION OF THE DRAWING
The above and other objects, features, and advantages will become more
readily apparent from the following description, reference being made to
the accompanying drawing in which:
FIG. 1a is a plan view of a UV detector according to the invention;
FIG. 1b is a cross sectional view through this detector;
FIG. 2 is a graph showing light intensity versus distance from the surface
exposed to radiation for three sample wavelengths;
FIG. 3 is a graph of quantum efficiency versus radiation wavelength for a
high selectivity detector according to the invention; and
FIG. 4 is a graph similar to FIG. 3 of a detector having high sensitivity
also in the visible range.
SPECIFIC DESCRIPTION
From FIG. 1a the various layers on the substrate 6 can be seen and the
model grid 5 is likewise visible. The substrate 6 is a transparent element
which can be glass, quartz, etc.
From FIG. 1b, it will be apparent that above the transparent conductor 4 on
the substrate 6 and between the transparent conductor 4 and the conductor
5 in the form of the model grid, a rectifying Junction is provided which
comprises, in succession the p.sup.+ amorphous hydrogenated and
carbon-containing silicon layer 2, the intrinsic amorphous hydrogenated
and carbon-containing silicon layer 1 and the n.sup.+ amorphous
hydrogenated silicon layer 3.
FIGS. 1a and 1b show an intermediate layer between two p.sup.+ and n.sup.+
amorphous silicon doped layers, for the purpose of forming the rectifying
junction. This intermediate layer must have a very slight defect state
this is obtained, for example, by using a low doping or undoped
(intrinsic) layer so as to form, respectively, a p.sup.+ -p.sup.- n.sup.-
-n.sup.+ or a p+-i-n.sup.+ structure. In the case of the p.sup.+
-i-n.sup.+ detector, the application of a reverse polarization may not be
effective, because the junction would be crossed by a significant dark
current due to microscopic short circuits which would penalize the
relationship between photocurrent and dark noise. The presence of a
significant inverse saturation current is typical of very thin and defect
p-i-n diodes in a Si:H. From these considerations, a better operation of
the device in question is to be expected in terms of S/N ratio, with
polarization voltages around zero.
Concerning FIG. 1b, it must be borne in mind that the real dimensions of
the layers will be qualitatively discussed below. In the drawing the
relationships between dimensions have not been maintained.
In observing FIG. 1a, it noted that the electromagnetic radiation
penetrates into the amorphous device through the open regions of the metal
grid and crosses serially the following layers (FIG. 1b): layer 2, p.sup.+
of a-SiC:H; layer 1, intrinsic or p-/n- of a-SiC:H; layer 3, n.sup.+ of
a-Si:H. The order of the various layers composing the device is reversible
in case of use of a substrate transparent to the UV light, such as quartz,
magnesium fluoride or the like. In this case, the structure becomes:
substrate, metal grid, p.sup.+ -n.sup.+ junction with the previously
described intermediate layer, rear electrode.
However, the first active layer is the p.sup.+ layer (which we will call
the window layer). Based on the typical values of the absorption
coefficients provided in the literature it is found that, in a p.sup.+
layer of a-SiC:H with 50% percent of carbon, all the ultraviolet radiation
is absorbed essentially in the first 5 nm (FIG. 2). This dimension is of
the same order as the diffusion length of the electrons (minority
carriers) in that material. Hence, by dimensioning the thickness of the
p.sup.+ layer around this value there is a good probability for collection
of the electron-hole pairs photogenerated in the p.sup.+ layer after the
absorption of the UV. Smaller thicknesses of the p.sup.+ layer are
technologically inadvisable as they worsen the quality of the junction
whereas greater a-SiC:H thicknesses would penalize collection of carriers
and, consequently, the photocurrent. Greater thicknesses could be
considered if the p.sup.+ layer of an a-SiC:H alloy less rich in carbon
could be considered, since the diffusion length is greater. In this case
the transparency to the visible would be less (and selectivity as well).
The high value of the energy gap of the p.sup.+ layer (also exceeding 2 eV)
enhances transmission of the visible radiation which, thus, in negligibly
absorbed in the p.sup.+ layer.
The holes photogenerated in the p.sup.+ by the UV light are collected by
the metal grid electrode. The electrons photogenerated in the p.sup.+
layer diffuse and reach the n.sup.+ layer, where they are the majority
carriers and, thus, are collected by the rear electrode. In the
intermediate layer the electrons move by effect of the external voltage or
of the contact potential, having specified that this photodetector may or
may not be polarized, depending on the structure chosen.
By fixing the thickness of the intermediate a-SiC:H layer as on the order
of some tens of nanometers the collection probability is high. The visible
radiation is absorbed in the intermediate region in an amount that
increases with the increase in thickness. With a thickness of some tens of
nanometers a significant absorption is obtained of the portion of the
visible spectrum adjacent to the ultraviolet radiation, i.e., of the blue
(while the red would be so weakly absorbed as to cross almost unhindered
through the a-SiC:H intermediate layer to arrive in layer n.sup.+ and to
escape through the transparent electrode and substrate).
The use of an intermediate layer of a-SiC:H decreases the absorption of the
visible radiation and contributes to the selectivity of the detector by
comparison with the case of an a-Si:H intermediate layer.
In the thick n.sup.+ layer of a-Si:H the collection of photogenerated
carriers is an event of negligible probability.
In other words, selectivity in the ultraviolet band is ensured by the
following factors:
transparency of the p.sup.+ layer to the visible light (a-SiC:H);
a low absorption profile in the visible range also in the intermediate
layer (a-SiC:H);
the small thickness of the intermediate layer.
The high quantum efficiency in the ultraviolet radiation is ensured by the
relationship between the open area and the opaque area in the metal grid
and between thickness of the p.sup.+ layer and length of diffusion of
electrons.
From what has been said, a maximum sensitivity of the detector is obtained
in the ultraviolet radiation which progressively decreases with the
increase in wavelength, i.e. from the blue toward the red, with
insignificant efficiency values in the infrared.
Let us clarify the meaning of maximum sensitivity in ultraviolet radiation.
In principle, if the diffusion length in the p.sup.+ layer was much
greater than its thickness, all the carriers photogenerated in the p.sup.+
would have a high probability of contributing to the photocurrent even
those generated at the surface.
From literature it is known that the diffusion length of electrons in that
material does not exceed a few nanometers. Thus, the sensitivity should
decrease with wavelengths in the far UV range. Another mechanism takes
place, which increases the collection probability of photocarriers with
decreasing wavelengths, namely, the electron yield effect. This effect
gives rise to the generation of more than one electron-hole couple after
absorption of one photon if the photon energy greatly exceeds the energy
gap of the semiconductor. For example, photons with a wavelength below 300
nm have energies greater than two times the energy gap of the a-SiC:H used
in the p.sup.+ a layer, and can contribute to a multiple photogeneration
mechanism. Considering all the effects mentioned above the sensitivity of
the photodetector should remain constant over the entire UV band. And also
in the far-ultraviolet and beyond.
FIG. 3 shows diagrammatically the expected quantum efficiency curve in
arbitrary units for a structure like that shown in FIG. 1, calculated on
the basis of the absorption relative to the various wavelengths and of the
typical values provided in the literature of the transport constants of
the semiconductors under study.
With this invention it is possible to produce also a photodetector with
high spectral response in the UV and also in the visible region (i.e. of
the type schematically represented in FIG. 4). In this case the structure
is the following: p.sup.+ -i-n.sup.+, modified according to the following
criteria:
an intrinsic layer is deposited in a-Si:H instead of a-SiC:H to increase
absorption of the visible;
its thickness is increased to a few hundred of nanometers as is used in
a-Si:H solar cells where it is intended to absorb the visible; and
a metal rear electrode is applied to reflect the still unabsorbed radiation
and cause it to pass again into the active layer.
The deposition technique used for the photodetector permits fabrication on
glass, plastic, metal and ceramic types of substrates (also opaque and
flexible substrates) on which a conductor material film has been
predeposited. In addition, this technique allows fabrication over any area
even very large area surfaces if the deposition machine is set for large
area work. Thus, by utilizing the concepts just expressed, large area,
two-dimensional matrixes can be fabricated of the photodetectors (pixels)
whose minimum dimension depends on external factors (the lithographic
technique in use) and determines the spatial resolution. To this is added
the fact that the thicknesses and profiles of absorption of the layers can
be optimized so as to make the window functioning in the ultraviolet
spectrum very selective. Furthermore, if the rear electrode and the
substrate are transparent to the visible, a large part of the transmitted
radiation can finally exit from the detector thereby enhancing
selectivity. The thicknesses and profiles of absorption of the layers can
be optimized so as to extend the operation of the detector to the visible.
As mentioned earlier, to further increase the efficiency of detection in
the visible, a metal rear electrode can be used which reflects radiation,
making it pass through the intrinsic layer for the second time.
The present structure can be lengthened to a back-to-back diode structure
(of the type: p.sup.+ -i-n.sup.+ -i-p.sup.+), still composed of thin films
of hydrogenated amorphous silicon and its alloys such that they present a
second operation band in another spectral range. The back-to-back diode
photodetector presents a second operation band at longer wavelengths,
which can be in the infrared region if the second added diode is made of
hydrogenated amorphous silicon-germanium alloy. The photodetector diode
structure can be further lengthened to a p.sup.+ -i-n.sup.+ -i-n.sup.+
-i-p.sup.+ structure (according to the information of Italian application
corresponding to copending U.S. application Ser. No. 08/437,498, now U.S.
Pat. No. 5,557,133). Also composed of thin films of hydrogenated amorphous
silicon and its alloys which has a second and third operation bands
centered at longer wavelengths under the polarization conditions explained
in the above-mentioned patent.
Top